THEORETICAL-ANALYSIS OF NMR EXPERIMENTS IN NORMAL AND SUPERCONDUCTINGSTATES OF HIGH-T-C SUPERCONDUCTORS

The Knight shift and T-1- and T-2-rates of YBa2Cu3O6+x in the normal a nd superconducting state are modeled by calculating the magnetic susce ptibility in the bi-layer Hubbard model within various approximations. hn optimal sit of parameters (OSP) is found in the RPA approximation which fits experiments oil YBCO for optimal and nearly optimal doping. The analysis of the self-consistent FLEX approximation for the partic le self-energy and susceptibility shows that the latter is renormalize d quantitatively but not qualitatively. The differences in the oxygen and copper T-1-rates are explained by using the OSP parameters and ass uming the finite hyperfine coupling C' between O-17 and next-nearest n eighboring Cu spins. The numerical analysis of T-1(-1) and T-2(-1) and the ratio T-63(1ab)-1/T-63(1c)-1 in the superconducting stale support s strongly the idea of ci-wave pairing in YBa2Cu3O7 with much stronger intraplane rather than interplane pairing. It is also shown that the simple RPA or FLEX approximations are inadequate in explaining NMR dat a in underdoped YBCO systems; where antiferromagnetic fluctuations are very pronounced. (C) 1998 Elsevier Science B.V.